umutarpayy/sosyal8_bert

This model is a fine-tuned version of dbmdz/bert-base-turkish-cased on an unknown dataset. It achieves the following results on the evaluation set:

• Train Loss: 0.0506
• Train Accuracy: 0.9878
• Validation Loss: 0.0372
• Validation Accuracy: 0.9858
• Epoch: 12

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': {'module': 'transformers.optimization_tf', 'class_name': 'WarmUp', 'config': {'initial_learning_rate': 3e-05, 'decay_schedule_fn': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 3e-05, 'decay_steps': 10776, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'warmup_steps': 1197, 'power': 1.0, 'name': None}, 'registered_name': 'WarmUp'}, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False}
• training_precision: float32

Training results

Train Loss	Train Accuracy	Validation Loss	Validation Accuracy	Epoch
2.7138	0.3147	1.2947	0.6764	0
1.1971	0.6880	0.8961	0.7619	1
0.9229	0.7514	0.7018	0.8148	2
0.7358	0.7940	0.5743	0.8406	3
0.5680	0.8434	0.3641	0.9023	4
0.4190	0.8844	0.2454	0.9328	5
0.2932	0.9188	0.1395	0.9607	6
0.2126	0.9406	0.1021	0.9715	7
0.1567	0.9560	0.0783	0.9749	8
0.1145	0.9684	0.0633	0.9817	9
0.0860	0.9767	0.0451	0.9885	10
0.0663	0.9826	0.0414	0.9864	11
0.0506	0.9878	0.0372	0.9858	12

Framework versions

• Transformers 4.50.0
• TensorFlow 2.18.0
• Datasets 3.5.0
• Tokenizers 0.21.1